The latest informe de Eye on the Market of JP Morgan delves into an analysis of the latest advances in artificial intelligence. It is, strictly speaking, a warning about a regime change.
The development of frontier models like Mythos marks a turning point: " AI stops being a productivity vector and becomes a systemic risk factor with direct implications for financial markets, banking stability, and global technological infrastructure", the document warns.
What matters is not only the power of the model. AI doesn't just look for flaws: it also looks for how to exploit them. According to the report, their abilities were not designed explicitly, but rather emerge as a byproduct of their reasoning capacity.
The report draws a disturbing parallel with the dynamics of nuclear proliferation: "An initial period of concentration of capabilities in a few actors followed by expansion toward a more multipolar world and, by definition, a more unstable one. As these tools spread, the possibility of large-scale cyber events ceases to be a tail-end scenario to become a factor to incorporate into valuaciones, spreads y primas de riesgo ".
AI puts finance at risk
In that context, the banking system appears as the most exposed link "due to structural dependence on shared infrastructure", since entities operate over an interconnected network of cloud providers, third-party software, payment systems and clearing networks.
That interdependence, which seeks efficiency, becomes a channel for risk transmission when the threat is capable of escalating simultaneously across multiple layers of the system: an isolated event can quickly transform into a systemic problem.
This redefines the regulatory agenda. Cybersecurity ceases to be an operational aspect to become a central dimension of financial stability. In practical terms, it implies greater supervisory requirements, pressure on the resilience of technology providers, and a structural increase in compliance costs.
Large tech companies, for their part, find themselves in an ambivalent position. They are, at the same time, the major beneficiaries of artificial intelligence development and the nucleus where risk is concentrated . The report highlights that many of the vulnerabilities detected affect basic components of the digital ecosystem, from operating systems to open source libraries, which makes technological infrastructure at a single point of failure with global reach.
To this is added a critical problem: " The gap between vulnerability detection and correction. In cloud environments, that process can be relatively agile, but in legacy systems, especially in industrial or outdated infrastructure, patching cycles can extend over years. That temporal gap is, in practice, the space where it is defined whether technology acts as a defense mechanism or as an attack tool".
The impact on the crypto ecosystem is more subtle, but no less relevant. Although the sector's narrative relies on decentralization, much of its operations depend on the same technological infrastructure as the traditional financial system, such as cloud providers, code libraries, and execution environments that are not immune.
The emergence of models capable of exploiting flaws automatically reduces security margins and raises operational risk, while opening the door to a new generation of AI-based audit and defense tools.
How artificial intelligence changes the global risk map
Néstor Markowicz, COO of CertiSur, says that warnings from organizations like the Federal Reserve, the U.S. Treasury, and financial players like JP Morgan reflect a change of scale in artificial intelligence.
The expert argues that it is not just a matter of more capacity, "but of greater autonomy and speed applied to critical environments", and asserts that it is a turning point that forces us to rethink cybersecurity at a systemic level, "especially in the protection of digital identities and transactions ".
Along those lines, he suggests that what is truly disruptive about these models is that they integrate capabilities that were previously separate: " They can detect vulnerabilities, analyze them, and generate ways to exploit them in very short timeframes ".
He asserts that this completely changes the equation, since tasks that previously took weeks or months are now performed automatically and at scale, which reinforces the role of robust cryptography, identity management, and digital certificates as a first line of defense.
Markowicz argues that the the most exposed sectors are those with high technological dependence and systemic impact : financial, critical infrastructure, energy and telecommunications, government agencies, health and large digital platforms.
"In these environments, it ensures that having solid authentication, digital signature, and certificate management schemes is not optional ", but rather a central part of operational resilience", the expert analyzes.
Furthermore, he warns that the main risk is the scalability of the attack, since artificial intelligence makes it possible to industrialize the discovery and exploitation of vulnerabilities , and adds that the democratization of cybercrime enables actors with less experience to execute sophisticated attacks.
He also raises a systemic risk, since the same model can identify common weaknesses across multiple organizations simultaneously , which is why it considers it key to strengthen the digital trust infrastructure to guarantee the integrity, authenticity, and confidentiality of information.
In this context, he confesses that the system enters a new race between offensive and defensive capabilities , where the same technology that can be used to attack is also fundamental to improve early detection and prevention. It notes that at CertiSur they work on that axis, through digital identity solutions, PKI and cryptographic protection.
Finally, he asserts that the path forward is not to stop the technology, but to accompany it with greater maturity in cybersecurity, regulation, and risk management . It notes that these advances also open an opportunity: using AI to anticipate threats and strengthen digital trust, where tools such as digital signatures, SSL/TLS certificates and identity management become key pillars in an increasingly automated environment.
AI accelerates risk: why attacking is now cheaper and faster
Facundo Balmaceda, a cybersecurity specialist at SONDA Argentina, says that the advance of artificial intelligence represents not only a technological improvement, "but a change of scale with direct impact on markets and in the dynamics of risk".
He argues that these increasingly powerful tools can analyze information, make decisions, and automate tasks at an unprecedented speed, which redefines both productivity and the attack surface.
In that sense, he asserts that the strength of AI lies in "freeing humans from repetitive tasks to focus them on strategic decisions". However, he suggests that this same advantage can be exploited for malicious purposes if adequate controls do not exist.
Balmaceda agrees that the the most exposed sectors are those that handle critical information or key infrastructure, such as financial, energy, industry, health, and public sectors . It explains that the problem is not AI itself, "but its capacity to amplify errors and accelerate attacks when implemented without a cybersecurity resilience strategy or comprehensive security".
In that framework, he highlights a central point: " Artificial intelligence drastically reduces the cost of attacks, allowing actors with few resources to scale offensives that previously required complex structures".
Furthermore, he argues that the real risk is not in the technology, but in its adoption without clear rules, without human oversight, and without protecting data privacy and integrity. He asserts that organizations that understand this change in time will gain competitiveness , while those that do not will be exposed to unnecessary risks.
Finally, he confesses that the current challenge " not to halt the advance of artificial intelligence, but to learn to govern it efficiently".